Self-repairing Learning Rule for Spiking Astrocyte-Neuron Networks

Junxiu Liu, Liam McDaid, Jim Harkin, John Wade, Shvan Karim, Anju P. Johnson, Alan G. Millard, David M. Halliday, Andy M. Tyrrell, Jon Timmis

Research output: Chapter in Book/Report/Conference proceedingChapter

2 Citations (Scopus)

Abstract

In this paper a novel self-repairing learning rule is proposed which is a combination of the spike-timing-dependent plasticity (STDP) and Bienenstock, Cooper, and Munro (BCM) learning rules: in the derivation of this rule account is taken of the coupling of GABA interneurons to the tripartite synapse. The rule modulates the plasticity level by shifting the plasticity window, associated with STDP, up and down the vertical axis as a function of postsynaptic neural activity. Specifically when neurons are inactive, the window is shifted up the vertical axis (open) and as the postsynaptic neuron activity increases and, as learning progresses, the plasticity window moves down the vertical axis until learning ceases. Simulation results are presented which show that the proposed approach can still maintain the network performance even with a fault density approaching 80% and because the rule is implemented using a minimal computational overhead it has potential for large scale spiking neural networks in hardware.
LanguageEnglish
Title of host publicationNeural Information Processing
Pages384-392
DOIs
Publication statusE-pub ahead of print - 26 Oct 2017

Fingerprint

Neurons
Plasticity
Network performance
Astrocytes
Neural networks
Hardware

Keywords

  • Astrocyte-neuron
  • network Learning window
  • Self-repair
  • Fault tolerance
  • adaptive
  • neural network

Cite this

Liu, J., McDaid, L., Harkin, J., Wade, J., Karim, S., Johnson, A. P., ... Timmis, J. (2017). Self-repairing Learning Rule for Spiking Astrocyte-Neuron Networks. In Neural Information Processing (pp. 384-392) https://doi.org/10.1007/978-3-319-70136-3_41
Liu, Junxiu ; McDaid, Liam ; Harkin, Jim ; Wade, John ; Karim, Shvan ; Johnson, Anju P. ; Millard, Alan G. ; Halliday, David M. ; Tyrrell, Andy M. ; Timmis, Jon. / Self-repairing Learning Rule for Spiking Astrocyte-Neuron Networks. Neural Information Processing. 2017. pp. 384-392
@inbook{7004b9adac31480fa075dba447b7c1c5,
title = "Self-repairing Learning Rule for Spiking Astrocyte-Neuron Networks",
abstract = "In this paper a novel self-repairing learning rule is proposed which is a combination of the spike-timing-dependent plasticity (STDP) and Bienenstock, Cooper, and Munro (BCM) learning rules: in the derivation of this rule account is taken of the coupling of GABA interneurons to the tripartite synapse. The rule modulates the plasticity level by shifting the plasticity window, associated with STDP, up and down the vertical axis as a function of postsynaptic neural activity. Specifically when neurons are inactive, the window is shifted up the vertical axis (open) and as the postsynaptic neuron activity increases and, as learning progresses, the plasticity window moves down the vertical axis until learning ceases. Simulation results are presented which show that the proposed approach can still maintain the network performance even with a fault density approaching 80{\%} and because the rule is implemented using a minimal computational overhead it has potential for large scale spiking neural networks in hardware.",
keywords = "Astrocyte-neuron, network Learning window, Self-repair, Fault tolerance, adaptive, neural network",
author = "Junxiu Liu and Liam McDaid and Jim Harkin and John Wade and Shvan Karim and Johnson, {Anju P.} and Millard, {Alan G.} and Halliday, {David M.} and Tyrrell, {Andy M.} and Jon Timmis",
year = "2017",
month = "10",
day = "26",
doi = "10.1007/978-3-319-70136-3_41",
language = "English",
isbn = "978-3-319-70095-3",
pages = "384--392",
booktitle = "Neural Information Processing",

}

Liu, J, McDaid, L, Harkin, J, Wade, J, Karim, S, Johnson, AP, Millard, AG, Halliday, DM, Tyrrell, AM & Timmis, J 2017, Self-repairing Learning Rule for Spiking Astrocyte-Neuron Networks. in Neural Information Processing. pp. 384-392. https://doi.org/10.1007/978-3-319-70136-3_41

Self-repairing Learning Rule for Spiking Astrocyte-Neuron Networks. / Liu, Junxiu; McDaid, Liam; Harkin, Jim; Wade, John; Karim, Shvan; Johnson, Anju P.; Millard, Alan G.; Halliday, David M.; Tyrrell, Andy M.; Timmis, Jon.

Neural Information Processing. 2017. p. 384-392.

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - Self-repairing Learning Rule for Spiking Astrocyte-Neuron Networks

AU - Liu, Junxiu

AU - McDaid, Liam

AU - Harkin, Jim

AU - Wade, John

AU - Karim, Shvan

AU - Johnson, Anju P.

AU - Millard, Alan G.

AU - Halliday, David M.

AU - Tyrrell, Andy M.

AU - Timmis, Jon

PY - 2017/10/26

Y1 - 2017/10/26

N2 - In this paper a novel self-repairing learning rule is proposed which is a combination of the spike-timing-dependent plasticity (STDP) and Bienenstock, Cooper, and Munro (BCM) learning rules: in the derivation of this rule account is taken of the coupling of GABA interneurons to the tripartite synapse. The rule modulates the plasticity level by shifting the plasticity window, associated with STDP, up and down the vertical axis as a function of postsynaptic neural activity. Specifically when neurons are inactive, the window is shifted up the vertical axis (open) and as the postsynaptic neuron activity increases and, as learning progresses, the plasticity window moves down the vertical axis until learning ceases. Simulation results are presented which show that the proposed approach can still maintain the network performance even with a fault density approaching 80% and because the rule is implemented using a minimal computational overhead it has potential for large scale spiking neural networks in hardware.

AB - In this paper a novel self-repairing learning rule is proposed which is a combination of the spike-timing-dependent plasticity (STDP) and Bienenstock, Cooper, and Munro (BCM) learning rules: in the derivation of this rule account is taken of the coupling of GABA interneurons to the tripartite synapse. The rule modulates the plasticity level by shifting the plasticity window, associated with STDP, up and down the vertical axis as a function of postsynaptic neural activity. Specifically when neurons are inactive, the window is shifted up the vertical axis (open) and as the postsynaptic neuron activity increases and, as learning progresses, the plasticity window moves down the vertical axis until learning ceases. Simulation results are presented which show that the proposed approach can still maintain the network performance even with a fault density approaching 80% and because the rule is implemented using a minimal computational overhead it has potential for large scale spiking neural networks in hardware.

KW - Astrocyte-neuron

KW - network Learning window

KW - Self-repair

KW - Fault tolerance

KW - adaptive

KW - neural network

U2 - 10.1007/978-3-319-70136-3_41

DO - 10.1007/978-3-319-70136-3_41

M3 - Chapter

SN - 978-3-319-70095-3

SP - 384

EP - 392

BT - Neural Information Processing

ER -